Thanks Johan - good advice as always I will look for the troublesome spot.
Yes on one side there are two small "limbs" between the shell and the core (sorry i don't know the correct terminology) which seems to 'fill the gap' and make it a little tighter. But on the other side there are no limbs. I only noticed this by comparing it to the output transformers, which have the limbs on both sides. But I can't tell from the Edcor site what the correct layout should be....
🙂
ooopppss Sorry,glad you found it ! My brain is running down,they have removed part of it already,😱 I copied it maybe too close,
NS
Ok just took some measurements at idle:
240v 60hz
0.55 amps, 98 watts, power factor 71.
240v 50hz
0.66 amps, 115 watts, power factor 72.
I'm not sure how the power factor measurement works, so I'll leave that one for the experts, but it does seem the extra current drawn by the 50hz is causing the grief.
240v 60hz
0.55 amps, 98 watts, power factor 71.
240v 50hz
0.66 amps, 115 watts, power factor 72.
I'm not sure how the power factor measurement works, so I'll leave that one for the experts, but it does seem the extra current drawn by the 50hz is causing the grief.
idle and unloaded?
what is the va rating of that traffo?
it looks to me that those figures are on the high side...
what is the va rating of that traffo?
it looks to me that those figures are on the high side...
That's also got the variac/frequency converter in series - the energy meter only plugs into the wall so we can subtract 30w which is drawn by the frequency converter at idle.
I just reduced the voltage down to 115v and switched between 50hz and 60hz and it is silent as a grave on both settings!
I just reduced the voltage down to 115v and switched between 50hz and 60hz and it is silent as a grave on both settings!
are you sure you got the primaries in series? 😉
one way to confirm is to inject 6 volts on a secondary winding and measure voltage at the primary side...
one way to confirm is to inject 6 volts on a secondary winding and measure voltage at the primary side...
Lordearl,
Never mind the power factor for now; that is a large increase from 60 Hz to 50 Hz .... something the matter. I presume you mean with the amplifier on and at idle. There should then only be a small increase in watt. (Typically some 15% of that, unless the transformer goes into unnatural sudden saturation.)
If you can still measure only the magnetising current/losses, i.e. with all tubes pulled including the rectifier, and hopefully no other power consuming elements.
I must bow out now and go to bed (already dawn here and up all night); will contemplate some more.
Never mind the power factor for now; that is a large increase from 60 Hz to 50 Hz .... something the matter. I presume you mean with the amplifier on and at idle. There should then only be a small increase in watt. (Typically some 15% of that, unless the transformer goes into unnatural sudden saturation.)
If you can still measure only the magnetising current/losses, i.e. with all tubes pulled including the rectifier, and hopefully no other power consuming elements.
I must bow out now and go to bed (already dawn here and up all night); will contemplate some more.
240v 60hz
0.55 amps, 98 watts, power factor 71.
240v 50hz
0.66 amps, 115 watts, power factor 72.
The current ratio is looking alright, but the current should be more like .055 Amp and .066 Amp.
If the meter is before the freq. converter, that is the problem, switching spikes from the converter messing with the meter. The meter must be at the xfmr input. The freq. converter most likely turns reactive power drawn by the xfmr into real power from the wall outlet besides. (no provision to send power backwards) The converter draws more internal power as well with a load, since it is probably only 60% efficient.
0.55 amps, 98 watts, power factor 71.
240v 50hz
0.66 amps, 115 watts, power factor 72.
The current ratio is looking alright, but the current should be more like .055 Amp and .066 Amp.
If the meter is before the freq. converter, that is the problem, switching spikes from the converter messing with the meter. The meter must be at the xfmr input. The freq. converter most likely turns reactive power drawn by the xfmr into real power from the wall outlet besides. (no provision to send power backwards) The converter draws more internal power as well with a load, since it is probably only 60% efficient.
Last edited:
Yes I had thought we need to put the meter between then variac and the xfmr but I don't have the correct adapter to do it.
Best I can do is provide measurements of the variac at idle, then under load:
Converter/variac alone at idle:
240v 60Hz: 0.06A, 11W
240v 50Hz: 0.06A, 11.6W
Amp connected to converter/variac without tubes:
240v 60Hz: 0.14A, 24.4W
240v 50Hz: 0.27A, 45.8W
Lordearl .....
something is not kosher here. You tempt me to go do my own tests before the year is out!
Easy this side: Drive a test transformer with another one, stepped up to 240V from my amplifier output, fed with 50 Hz and 60 Hz signals. It will be 50 Hz transformers; can then go down to say 42 Hz and see.
Not sure what your transformer is to start with: Some 200VA? Would expect losses to be in the 20W area then .... or perhaps have my own horizons broadened .. 😱
something is not kosher here. You tempt me to go do my own tests before the year is out!
Easy this side: Drive a test transformer with another one, stepped up to 240V from my amplifier output, fed with 50 Hz and 60 Hz signals. It will be 50 Hz transformers; can then go down to say 42 Hz and see.
Not sure what your transformer is to start with: Some 200VA? Would expect losses to be in the 20W area then .... or perhaps have my own horizons broadened .. 😱
This really seems to indicate insufficient inductance for 50Hz/240V operation, all of that additional power (17W) will be dissipated as heat in the windings of that transformer. What is your line voltage? I would have expected just a few more watts switching from 60Hz to 50Hz.. (Looks like core losses to me given the idle consumption is not greatly higher.)
Amp connected to converter/variac without tubes:
240v 60Hz: 0.14A, 24.4W
240v 50Hz: 0.27A, 45.8W
Try the amp connected, but with 50/60 (240) = 200 V at 50 Hz
That should give the same magnetic "saturation" level in the xfmr as the 240V 60 Hz.
If the current drops back to 0.14A there, then the xfmr can't handle the 50 Hz.
If your variac allows V boost mode, you could also try 60/50 (240) = 288 V, on 60 Hz, to see if the current goes up to 0.27 A. (assuming no electrolytic caps on the xfmr output that would be V stressed)
240v 60Hz: 0.14A, 24.4W
240v 50Hz: 0.27A, 45.8W
Try the amp connected, but with 50/60 (240) = 200 V at 50 Hz
That should give the same magnetic "saturation" level in the xfmr as the 240V 60 Hz.
If the current drops back to 0.14A there, then the xfmr can't handle the 50 Hz.
If your variac allows V boost mode, you could also try 60/50 (240) = 288 V, on 60 Hz, to see if the current goes up to 0.27 A. (assuming no electrolytic caps on the xfmr output that would be V stressed)
Last edited:
OCH! Me and me big mouth 
I did the above tests, using a power transformer (I thought) had moderately good iron.
Not so.
Working at the design voltage, which should just begin to be on the 'decline' of the B-H slope, I had my horizons broadened. The best way to represent the results is to normalise to the loss at 50 Hz:
I found that the loss at 60 Hz was 0,4 of that, and at 42 Hz ( the ratio of 50/60 Hz) 2,6 times the 50 Hz value! So yes, depending on the design, there apparently will be a measureable difference going from 60 - 50 Hz. (Disillusioned by this I left the test transformer on for about an hour; slight rise in temperature, not abnormal for normal transformer heat increase from iron losses.)
THUS: Unfortunately the OP might well be burdened with some core saturation. I might only hope that you do find the source of the hum .....
And a more pleasant task remains for me: The very best wishes for 2015 to all, thanks for patience and presenting a real good forum (but you don't need me to tell you that). I am honoured to be a member here.
I did the above tests, using a power transformer (I thought) had moderately good iron.
Not so.
Working at the design voltage, which should just begin to be on the 'decline' of the B-H slope, I had my horizons broadened. The best way to represent the results is to normalise to the loss at 50 Hz:
I found that the loss at 60 Hz was 0,4 of that, and at 42 Hz ( the ratio of 50/60 Hz) 2,6 times the 50 Hz value! So yes, depending on the design, there apparently will be a measureable difference going from 60 - 50 Hz. (Disillusioned by this I left the test transformer on for about an hour; slight rise in temperature, not abnormal for normal transformer heat increase from iron losses.)
THUS: Unfortunately the OP might well be burdened with some core saturation. I might only hope that you do find the source of the hum .....
And a more pleasant task remains for me: The very best wishes for 2015 to all, thanks for patience and presenting a real good forum (but you don't need me to tell you that). I am honoured to be a member here.
In the "linear" range, the xfmr core losses typically scale as the square of the magnetic flux level (ie, Volts). So (60/50)^2 = 1.44 and (50/60)^2 = 0.7 would be the expected, minimum.
No doubt the factors will be bigger near saturation. So the 2X factor seen for the xfmr being scrutinized (which is better than the 2.6 just measured by Johan) looks to be fine.
Probably just something loose in the xfmr that is buzzing with stronger excitation.
No doubt the factors will be bigger near saturation. So the 2X factor seen for the xfmr being scrutinized (which is better than the 2.6 just measured by Johan) looks to be fine.
Probably just something loose in the xfmr that is buzzing with stronger excitation.
When you used another transformer in front of the noisy one it delivered a sine wave to the second one that had equal energy around zero volts. This is often not the case with the power line as there are devices that draw more or less power from the positive swing than the negative. This is often called DC on your AC. That offset will cause a buzzing noise from the core of some transformers.
The solution is to put a capacitor in series with the AC line. At .27 amps from a 240 volt line that is a load of 889 ohms. So a capacitor rated at 35 uF or more in series with the transformer's primary will solve the problem. There are motor run capacitors rated for 440 volts that would work and around here cost $25.00.
Most folks use a lower voltage capacitor (Often back to back electrolytics or AC rated ones) and place diodes across them to limit the voltage. If you go that route use three diodes in series and parallel that with another three facing the other way and use say 470 uF at 15V for the capacitors.
That should solve you problem.
The other technique is to actually step the voltage down a bit with say a 12 volt buck transformer.
I'll re-wire the xfmr for 120v primary and use the step down to avoid hum while I wait for Edcor to respond to my emails. It suddenly struck me that it's quite poor for the buzzing to be so different when alternating from 60 to 50 Hz so it surely is a manufacturing problem they should know about!
Not sure if it is dc on the mains as wouldn't that manifest itself in other transformers in the audio system? Is it possible to measure for dc on the mains? If so, what is an acceptable level of dc voltage?
Not sure if it is dc on the mains as wouldn't that manifest itself in other transformers in the audio system? Is it possible to measure for dc on the mains? If so, what is an acceptable level of dc voltage?
Yes you can measure for DC on the mains. A good digital scope will show it. You can build a gizmo to show it on a meter.
The reason why the transformer buzzes is that at 60 hertz or even a slightly lower voltage the core doesn't saturate as much. Other folks probably use a bit more iron in the core and can handle a bit more.
The reason why the transformer buzzes is that at 60 hertz or even a slightly lower voltage the core doesn't saturate as much. Other folks probably use a bit more iron in the core and can handle a bit more.
Hi,🙂
I copy and pasted and left the l off the end,sorry,
Try this one
Practical Transformer Winding
Thanks,😀
NS
- Status
- Not open for further replies.